Abstract Background and Aims ANCA-associated vasculitis (AAV) with glomerulonephritis (AAV-GN) still carries poor prognosis despite treatment improvement over the last years. Currently, improving assessment of disease activity and severity and defining risk factors for ESKD pose challenges. In this study we aim to determine and characterize risk factors associated with progression to ESKD or death in patients with AAV-GN. Method We retrospectively analysed all AAV-GN cases of our Center. ESKD was considered as a composite outcome of eGFR<15 or dialysis dependency. Biopsies were classified according to Berden's classification (BC), Brix's renal risk score (RRS), Mayo Clinic Chronicity Score (MCCS) and AAV-GN combined score (AAV-GN-CS). Categorical variables were analysed by non-parametric chi-square or fisher's exact test; and continuous variables by Mann-Whitney U test. Cox regression models were applied for death and Kaplan-Meier estimator. Significance value α = 0.05 was considered. Results A total of 74 patients, 56.8% male, 69 years-old median age were included. Of them, 81.2% were MPO, 11.6% PR3 and 7.2% negative. At presentation, 35.3% had alveolar haemorrhage, 13.4% GI and 66.7% constitutional manifestations. At diagnosis, median eGFR was 7.5 mL/min/1.73 m2 (5.0-13.0), proteinuria 2.2g/g (0.45-4.17) and all patients had haematuria. Initial median ANCA titer (ATi) was 100 UI/mL (51-134) and at last follow-up 44 (15-101). 64.2% of all patients required dialysis at presentation. Of these, 14.0% recovered until discharge. We also analyzed biopsies of 57 patients. According to BC, 15.8% were focal, 40.4% crescentic, 26.3% mixed, and 17.5% focal. ESKD risk was medium in 28.3% and high in 69.8%, as determined by RRS; whereas low in 31%, intermediate in 42.3% and high in 19% as per AAV-GN-CS. Induction treatment (IT) used was cyclophosphamide (CYC) in 46.3%, CYC plus rituximab (RTX) in 18.3% and RTX in 3.3%. Only 17 patients (26.2%) were added plasmapheresis (PLEX). Only 28% achieved kidney remission (KRm), 7% relapsed and 68.7% developed ESKD. Of them, 95.5% had eGFR<15 at diagnosis, 30% low C3, 37.5% crescentic pattern, 28.1% mixed and 28.1% sclerotic. 87% of ESKD patients had <25% normal glomeruli (NG), and 87.1% high ESKD risk (RRS). Progression to ESKD was associated with lower eGFR (p < 0.001) and low C3 at diagnosis (p = 0.026), lower %NG (p = 0.004), higher %CG (p = 0.004), higher %IFTA (p = 0.011), worse BC (p = 0.012), higher RRS (p < 0.001) and MCCS categories (p = 0.010), but not with AAV-GN-CS. ATi reduction seems to determine better renal outcome (p = 0.042). No significant differences were seen with IT or maintenance treatment, nor with PLEX addition. No patient treated with RTX, or combined treatment developed ESKD. Mortality was 32.8% and the best predictive model includes age, GI manifestations and KRm (p < 0.001). Age (p = 0.004), lower eGFR at diagnosis (p = 0.016) and GI manifestations (p = 0.003) were predictive of death, while achieving KRm improved survival (p = 0.009). Conclusion Our results reveal that AAV-GN patients have poor kidney outcome when presenting with severe kidney dysfunction, high ANCA titers and low C3 levels. Worse histological scoring is determinant of ESKD. The prediction power of AAV-GN-CS must be demonstrated in large cohorts. RTX effectivity on kidney survival must be further supported. Patient survival is related with younger age at diagnosis, organ systems involved at clinical presentation, and achievement of kidney remission.